HVAC monitoring method and apparatus

Abstract

Methods and devices for detecting inefficiencies in environmental control systems and for building a model for detecting the same. The model building method may comprise: receiving training data sets for one or more environmental control systems, identifying a category for each training data set indicative of the efficiency of the environmental control system over the training time period, computing one or more numerical features for each training data set based on setpoint data and sensor data, determining a measure of dependence between the identified category and the computed one or more numerical features, and creating a model based on the determined measure of dependence. The detection method may involve receiving a monitored data set, computing numerical features for the monitored data, and assigning a category indicative of the efficiency of a monitored environmental control system based on the model and the numerical features.

What is claimed is: 
     
       1. A computer-implemented method of controlling a monitored environmental control system arranged to control at least one environmental characteristic at a premises, the method comprising:
 receiving a monitored data set comprising setpoint data, control signal data for one or more appliances, and sensor data for an environmental characteristic over a monitoring time period, wherein the setpoint data specifies one or more target values for the environmental characteristic over the monitoring time period, and the sensor data comprises measured values for said environmental characteristic over the monitoring time period; 
 computing one or more numerical features for the monitored data set based on the setpoint data and sensor data for the environmental characteristic over the monitoring time period, wherein the numerical features comprise a mean activation period for the one or more appliances and a mean difference between the setpoint data and the sensor data; 
 receiving a model for detecting inefficiencies in environmental control systems, the model relating efficiency of environmental control systems to one or more numerical features derivable from a data set comprising setpoint data, control signal data for one or more appliances, and sensor data for an environmental characteristic controlled by environmental control systems, the one or more numerical features derivable from the data set comprising a mean activation period for the one or more appliances and a mean difference between the setpoint data and the sensor data, wherein the model comprises one or more rules for assigning one of a plurality of efficiency categories indicative of the efficiency of the monitored environmental control system based on the one or more numerical features, each rule comprising a threshold value for one of the one or more numerical features; 
 assigning an efficiency category indicative of the efficiency of the monitored environmental control system over the monitoring time period comprising inputting to the received model the one or more numerical features of the monitored data set computed based on the setpoint data, the control signal data for one or more appliances and the sensor data for said environmental characteristic over the monitoring time period; 
 identifying a solution based on the assigned efficiency category, wherein the solution comprises adjusting one or more settings of the environmental control system; and 
 sending a command to the environmental control system to adjust the one or more settings in response to identifying the solution. 
 
     
     
       2. A method according to  claim 1 , wherein each data set further comprises external environmental characteristic data relating to the environmental characteristic external to the premises; wherein at least one of the one or more numerical features is based on the external environmental characteristic data. 
     
     
       3. A method according to  claim 1 , wherein each data set further comprises control signal data for one or more appliances of the environmental control system, and wherein at least one of the one or more numerical features is based on the control signal data. 
     
     
       4. A method according to  claim 1 , further comprising:
 receiving a time series of setpoint values and sensor values for an extended time period; and dividing the extended time period into one or more regular, predetermined time slots to develop one or more data sets of setpoint values and sensor values, each data set corresponding to one of the time slots. 
 
     
     
       5. A method according to  claim 1 , wherein computing one or more numerical features for a data set based on the setpoint data and sensor data comprises:
 identifying one or more activation periods of appliances of the environmental control system for the monitoring or training period; computing the length of each of the identified one or more activation periods; and 
 using the computed length of each of the activation periods to compute the one or more numerical features for the data set. 
 
     
     
       6. A method according to  claim 1 , wherein computing one or more numerical features for a data set based on the setpoint data and sensor data comprises:
 computing the difference between the setpoint for an environmental characteristic and the sensed value of the environmental characteristic during activation periods. 
 
     
     
       7. A method according to  claim 1 , wherein the one or more numerical features comprise one or more of:
 the largest, mean and median values of the length of one or more activation periods of appliances of the environmental control system in the monitoring or training period; and a frequency or rate of activation periods of appliances of the environmental control system in the monitoring or training period. 
 
     
     
       8. A method according to  claim 1 , wherein the model comprises one or more rules for assigning an efficiency category indicative of the efficiency of the monitored environmental control system based on the one or more numerical features, each rule comprising a threshold value for one of the one or more numerical features, wherein each rule provides a first efficiency category to assign, or first possible range of efficiency categories to assign, when the numerical feature is below the threshold value and a second efficiency category to assign, or second possible range of efficiency categories to assign, when the numerical feature is above the threshold value. 
     
     
       9. A method according to  claim 1 , wherein computing one or more numerical features comprises computing at least two numerical features and wherein the model comprises two or more rules for assigning an efficiency category indicative of the efficiency of the monitored environmental control system based on the numerical features. 
     
     
       10. A method according to  claim 1 , wherein computing one or more numerical features comprises computing at least two numerical features and wherein the model comprises a weighting for each of the numerical features for assigning the efficiency category. 
     
     
       11. A method according to  claim 1 , wherein the one or more settings comprises a setpoint for the environmental control system, the method further comprising: in response to identifying the solution, providing a message to a user, the message comprising a selectable option to adjust the setpoint; and in response to receiving a user selection of the option to adjust the setpoint, sending a command to the environmental control system to adjust the setpoint. 
     
     
       12. A method according to  claim 1 , further comprising:
 alerting a user associated with the environmental control system in response to assigning an efficiency category indicative of an inefficient system, the alerting comprising one or more of: 
 displaying an alert on a display associated with the environmental control system; and transmitting an alert message to a user device associated with the user, the user device being a personal or mobile computing or communications device. 
 
     
     
       13. A method according to  claim 1 , comprising, in response to assigning an efficiency category indicative of an inefficient system, initiating a test procedure for testing the environmental control system, wherein the test procedure comprises:
 activating the environmental control system, wherein the activating overrides a control schedule active for the monitored environmental control system; measuring changes in the environmental characteristic in response to activation of the environmental control system; and performing further analysis of measured sensor data relating to the environmental characteristic to identify, quantify and/or classify the underperformance condition. 
 
     
     
       14. A method according to  claim 1 , further comprising:
 suppressing or filtering one or more alerts about the monitored environmental control system in response to assigning an efficiency category indicative of an inefficient system. 
 
     
     
       15. A method according to  claim 1 , wherein the model is built by a method comprising the steps of:
 receiving a plurality of training data sets for one or more environmental control systems, each training data set comprising setpoint data and sensor data for an environmental characteristic at a premises controlled by an environmental control system over a training time period; 
 identifying an efficiency category for each training data set indicative of the efficiency of the environmental control system over the training time period; computing one or more numerical features for each training data set based on the setpoint data and sensor data determining a measure of dependence between the identified efficiency category and the computed one or more numerical features; and creating a model based on the determined measure of dependence, wherein the model is configured to assign an efficiency category indicative of the efficiency of a monitored environmental control system based on a monitored data set comprising setpoint data and sensor data for an environmental characteristic at the premises controlled by the monitored environmental control system over a monitoring time period. 
 
     
     
       16. A method according to  claim 15 , wherein the step of identifying an efficiency category comprises outputting to a human operator a visual representation of each of the training data sets; and receiving an input from the human operator indicative of the efficiency category. 
     
     
       17. A method according to  claim 15 , wherein the step of identifying an efficiency category comprises receiving diagnostic data relating to the operation of one or more appliances in the environmental control system during the training time period. 
     
     
       18. A method according to  claim 1 , wherein computing one or more numerical features comprises detecting one or more of: a start time and an end time of one or more activation periods of appliances of the environmental control system for the monitoring time period in dependence on a difference between the setpoint data and the sensor data of the environmental characteristic. 
     
     
       19. A method according to  claim 18 , wherein the detecting comprises at least one of: inferring the start of the activation period in response to detecting one of the measured values not meeting the corresponding target value; and inferring the end of the activation period in response to detecting one of the measured values meeting the corresponding target value. 
     
     
       20. Apparatus for controlling an environmental control system arranged to control at least one environmental characteristic at a premises, comprising a processor and associated memory configured to:
 receive a monitored data set comprising setpoint data, control signal data for one or more appliances, and sensor data for an environmental characteristic over a monitoring time period, wherein: 
 the setpoint data specifies one or more target values for the environmental characteristic over the monitoring time period, and the sensor data comprises measured values for said environmental characteristic over the monitoring time period; 
 compute one or more numerical features for the monitored data set based on the setpoint data and sensor data for the environmental characteristic over the monitoring time period, wherein the numerical features comprise a mean activation period for the one or more appliances and a mean difference between the setpoint data and the sensor data; 
 receive a model for detecting inefficiencies in environmental control systems, the model relating efficiency of environmental control systems to one or more numerical features derivable from a data set comprising setpoint data, control signal data for one or more appliances, and sensor data for an environmental characteristic controlled by environmental control systems, the one or more numerical features derivable from the data set comprising a mean activation period for the one or more appliances and a mean difference between the setpoint data and the sensor data, wherein the model comprises one or more rules for assigning one of a plurality of efficiency categories indicative of the efficiency of the monitored environmental control system based on the one or more numerical features, each rule comprising a threshold value for one of the one or more numerical features; 
 assign an efficiency category indicative of the efficiency of the monitored environmental control system over the monitoring time period, comprising inputting to the received model the one or more numerical features of the monitored data set computed based on the setpoint data, the control signal data for one or more appliances, and sensor data for said environmental characteristic over the monitoring time period; 
 identify a solution based on the assigned efficiency category, wherein the solution comprises adjusting one or more settings of the environmental control system; and 
 send a command to the environmental control system to adjust the one or more settings in response to identifying the solution. 
 
     
     
       21. A non-transient computer-readable medium comprising instructions which, when executed by a computer, cause the computer to carry out a method of controlling an environmental control system arranged to control at least one environmental characteristic at a premises, the instructions configured when executed to:
 receive a monitored data set comprising setpoint data, control signal data for one or more appliances, and sensor data for an environmental characteristic over a monitoring time period, wherein the setpoint data specifies one or more target values for the environmental characteristic over the monitoring time period, and the sensor data comprise measured values for said environmental characteristic over the monitoring time period; 
 compute one or more numerical features for the monitored data set based on the setpoint data and sensor data for the environmental characteristic over the monitoring time period, wherein the numerical features comprise a mean activation period for the one or more appliances and a mean difference between the setpoint data and the sensor data; 
 receive a model for detecting inefficiencies in environmental control systems, the model relating efficiency of environmental control systems to one or more numerical features derivable from a data set comprising setpoint data, control signal data for one or more appliances, and sensor data for an environmental characteristic controlled by environmental control systems, the one or more numerical features derivable from the data set comprising a mean activation period for the one or more appliances and a mean difference between the setpoint data and the sensor data, wherein the model comprises one or more rules for assigning one of a plurality of efficiency categories indicative of the efficiency of the monitored environmental control system based on the numerical features, each rule comprising a threshold value for one of the one or more numerical features; 
 assign an efficiency category indicative of the efficiency of the monitored environmental control system over the monitoring time period, comprising inputting to the received model the one or more numerical features of the monitored data set computed based on the setpoint data, the control signal data for one or more appliances and the sensor data for said environmental characteristic over the monitoring time period; 
 identify a solution based on the assigned efficiency category, wherein the solution comprises adjusting one or more settings of the environmental control system; and 
 send a command to the environmental control system to adjust the one or more settings in response to identifying the solution.